Journal Club: Statistical shape models of male and female cochleae (and the rest of inner ear) from 54 human cadaveric temporal bones.

on

Today's journal article

Spedaliere C, Vaupotic A, Hwang J, Husein K, Hafidh M, Rioux K, Rohani SA, Agrawal SK, Ladak HM. Statistical shape modelling of the human inner ear through micro-computed tomography imaging. 

Why I picked this article

Our inner ear organ for hearing in mammals, the cochlea, has a beautiful and sophisticated structure - it spirals! It's a beautiful organ, but with its shape having references to its function. The outside end of the spiral is where the highest frequency sounds are detected, while the innermost tip of the spiral is where the lowest frequency sounds are detected. The 3D anatomy of the cochlea, however, also has variability between people. The variability of the cochlea and knowing the right "geometry range" is very important if we are to predict drug delivery and dispersion in the cochlea, or if we are to develop a new therapeutic approach. 

Statistical shape models (SSMs) can help us understand the anatomical variation between people by capturing detailed shape differences within a sample, and help us understand what the average cochlea looks like, but also where the variability tends to be higher within the structure. In this research, researchers have used SSMs to identify the normative range for inner anatomy in both 2D and 3D and provide reference points for the population average. 

Some of the research findings

Human temporal bone data: 
  • A total of 54 cadaveric human temporal bones were analysed from 41 donors. 
  • 15 male donors, 15 female donors, and 11 donors of unknown sex.
  • The age ranged from 57 to 94 years
  • micro-CT imaging (GE eXplore speCZT scanner), voxel size 50 μm, with a tube voltage and current of 90 kV and 40 mA.
Data analysis: 
  • Semi-automatic segmentation in 3D slicer using threshold painting and seed grow techniques. 
  • Segmentations were converted to 3D surface mesh model to check. 
  • Cochleae were aligned to a standard cochlear coordinate system (Verbist et al. 2010). 
  • Models were aligned to a standard scan with fiducial-based technique.
  • Models were registered to an arbitrary right inner ear using SlicerIGT module (ngi et al. 2016). 
  • Surface models were imported into Geomagic Studio 12 (3D systems Inc, 2014; http://www.
  • geomagic.com)
  • Within 3D slicer, fiducial markers were used on landmarks, placed on the 3D surface models, to make measurements. 
  • The volume was measured using 3D slicer. 
  • SSMs: aligned surface meshes were analysed using Shape-Works Studio (Cates et al., 2017; http://www.sci.utah.edu/software/shapeworks.html/)
  • Statistical analysis was conducted with Real Statistics Resource Pack software (Zaiontz, Release 8.9.1; https://real-statistics.com/).
Key findings: 
  • Normative linear and volumetric ranges consistent with prior reports; the comprehensive average geometrical parameters were generated for both the female and male populations.

Part of Figure 8d. The heat map showing deviations from the "mean cochlea", red colour shows higher deviation (0.6-0.7 mm) while blue colour shows low deviation (0.1-0.2 mm) from the mean. Carmine et al. 2025
Some key parameters:
  • The inner ear volumes were: 213.4 ± 19.1 for males, 182.3 ± 18.2 for females.
  • The cochlear volumes were: 86.9 ± 9.8 for males, 71.8 ± 8.3 for females.
  • The vestibular volumes were: 125.5 ± 14.5 for males, 109.1 ± 12.4 for females. 
Where the variation lives: 
  • Across the whole cohort, the parts that contributed the most to shape variabilities were: 
    • the semicircular canals.
    • the cochlear basal turn.
    • the cochlear hook region. 
  • Sex-specific patterns differed: 
    • Overall smaller cochlea /inner ear in females. 
    • The greater vestibular variability in females.
    • The greater basal-turn/hook variability in males.

Haruna's takeaway

What a comprehensive average data generation for the cochlear anatomy; this is such a useful paper for those who want to build a simple cochlear model, based on statistical averages. This is a type of information very useful for teaching and textbooks as well. 

To be honest, the variability between males and females was more than I had expected. Cochlea, but also the semi-circular canal length, seem quite different. The implication of the shape /size variability in hearing function is not known, so we cannot say much about the impact of the difference; however, it is a nice reminder that everyone's ears are very different. I also wonder if the average smaller volume of the inner ear/cochlea in females could impact the higher incidence of fluid-based diseases of the inner ear, like Meniere's disease, which is typically more common in females. Maybe less volume, less buffer for any excess volume/hydrops? 

 ------- 

This is Haruna's 93/100 of the 100-day challenge to post a science blog article every day! I love inner ear biology & cochlear physiology.